The present disclosure relates to a radiographic image analyzing device, a radiographic image analyzing method, and a radiographic image analyzing program for analyzing a radiographic image obtained by imaging a subject, and in particular to a radiographic image analyzing device, a radiographic image analyzing method, and a radiographic image analyzing program for analyzing a radiographic image obtained by imaging a subject to estimate the thickness of the subject at each position on the radiographic image.
It is conventionally known that, when a radiographic image of a subject is taken with radiation transmitted through the subject, image quality of the obtained radiographic image varies due to influences of scattering of the radiation in the subject, lowered transmittance of the radiation, etc., which influences are larger when the thickness of the subject is larger. To address this problem, techniques have been proposed to roughly estimate the thickness of the subject based on imaging conditions and various information, such as signal values of the radiographic image, the histogram width of signal values of the radiographic image, the length of the subject in the subject image in a predetermined direction, etc., and to change an image processing condition, such as scattered ray removal, performed on the obtained radiographic image or an imaging condition applied to radiographic imaging depending on the estimated thickness of the subject.
For example, a technique disclosed in Japanese Unexamined Patent Publication No. 2(1990)-244881 (hereinafter, Patent Document 1) involves: measuring pixel values of images obtained by radiographic imaging of a simulation subject having known thicknesses under known imaging conditions to prepare a lookup table that associates body thicknesses with pixel values; estimating a schematic body thickness distribution from pixel values of a subject image based on the lookup table; estimating a scattered ray component in the subject image based on the body thickness distribution of the subject image; and obtaining a processed image by subtracting the scattered ray component from the subject image.
D. E. G Trotter et. al., “Thickness-dependent Scatter Correction Algorithm for Digital Mammography”, Proc. SPIE, Vol. 4682, pp. 469-478, 2002, (hereinafter, Non-Patent Document 1) discloses a technique to estimate and remove a scattered ray component in a radiographic image based on the body thickness distribution of a human body. According to the image processing method of Non-Patent Document 1, a predetermined function is applied to an inputted subject image based on a body thickness distribution estimated from pixel values of the subject image to generate an estimated scattered ray image, which is an estimate of an image of scattered rays included in the subject image, and the estimated scattered ray image is subtracted from the subject image to generate an estimated primary ray image, which is an estimate of a primary ray image, from the inputted subject image. Then, operations to generate a new estimated scattered ray image by applying the predetermined function to the generated estimated primary ray image, and to generate a new estimated primary ray image by subtracting the new estimated scattered ray image from the subject image are iteratively performed until a certain convergence condition is satisfied to calculate a converged estimated scattered ray image, and the converged estimated scattered ray image is subtracted from the subject image to thereby obtain a processed image from which the scattered ray component is removed. Non-Patent Document 1 also disclose a method for adjusting the predetermined function used to estimate the image of scattered rays included in the subject image based on the body thickness.